Authorship

From time to time I have to go into our store to hunt through old (pre-war) reprints of medical research articles and I am always struck by the prevalence of single authorship in articles of that period. Single authorship in research articles is a rarity these days, and even review articles may have three or four authors. According to the National Library of Medicine, in 1950 the average number of authors per paper in Medline was 1.5; in 2011 it was 5. Pre-1975 the maximum number of authors on a paper in Medline was 37; by 2011 the maximum number was 3,172.

I think this is partly because research support staff have higher status (and probably more skills) nowadays than they did 60 years ago.  They are more likely to assert their rights when it comes to authorship credit. But it also reflects the fact that research has become more of a team effort, and those teams are getting larger.

What does it mean to be in the author list, or to put it another way, how do you qualify to be listed as an author? This is something I have been pondering as I have taken part in a working group charged with drawing up local guidelines, to try and ensure consistent practice across all labs at the Institute.

My contribution was to review what other guidelines say. The MRC recently published an updated version of its Good research practice: Principles and guidelines.  On authorship, it says:

Authorship should include all individuals who have made a substantial intellectual contribution and all authors are expected to take public responsibility for their contribution to the work. The MRC endorses the guidance of the Committee on Publication Ethics and the International Committee of Medical Journal Editors (ICMJE). The MRC endorses the ICMJE guidelines on authorship and contributorship; the practice of ‘honorary authorship’ is not acceptable. All contributions to the research must be clearly acknowledged and appropriate permissions sought for the use of the work of others. No person who fulfils the criteria for authorship should be excluded.

As I worked through guidelines from other bodies and from various journals I found that pretty well all of them also refer to the ICMJE guidelines.  These seem to encapsulate all that needs to be said on authorship. This was a relief as it meant I didn’t need to digest 94 different sets of guidelines, just one or two. (Wikipedia lists some authorship guidelines in other disciplines).

The ICMJE first drew up its “Uniform requirements for manuscripts submitted to biomedical journals” in 1978. Editors representing 19 medical journals drew up the guidelines, largely concerned with the structure and appearance of the articles, and practical issues (e.g. “Mail manuscripts in a heavy paper envelope, enclosing the manuscript and figures in cardboard, if necessary, to prevent bending of photographs during mail handling.”).  These guidelines (called the URM for short) were revised every few years. Their rules on formatting of bibliographies were very useful as we moved into the era of citation software in the 1980s and 1990s. The guidelines are now followed by a large number of journals.

Authorship was barely mentioned in the 1978 guidelines. In their history of the first 25 years of the URM, Edward Huth and Kathleen Case say:

The first and second editions of the URM mentioned authorship only briefly: “Acknowledge only persons who have made substantive contributions to the study.” By the third edition (1988) enough scandals had surfaced to lead the committee to define legitimate authorship in more detail. The key statement was, “Each author should have participated sufficiently in the work to take public responsibility for the content”, a criterion based on the view of Richard Hewitt, director of the Section of Publications of Mayo Clinic. This key statement was followed by more-specific criteria. In 1991, a statement was issued to cover
order of authorship.

The essence of the current ICMJE authorship guidelines are straightforward:

• Authorship credit should be based on
  1. Substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data;
  2. Drafting the article or revising it critically for important intellectual content;
  3. Final approval of the version to be published.

  Authors should meet conditions 1, 2, and 3.

• All persons designated as authors should qualify for authorship, and all those who qualify should be listed.
• Each author should have participated sufficiently in the work to take public responsibility for appropriate portions of the content.

In practice there is a good deal of subjective judgement involved in assessing who should or should not be listed as an author, so our local working group came up with a series of examples to aid the decision making process, listing the kinds of contributions that would or would not qualify for authorship, or acknowledgement.

These local guidelines have been endorsed by the Institute Director and have been disseminated to all staff. No doubt, just like the ICMJE guidelines, it will be necessary to keep them under review as research changes and author lists become still longer.

One area that we didn’t touch on is the issue that Jenny highlighted recently about the criteria for inclusion of the lab head, and their position as senior author.  I have seen examples of papers where the lab head does not appear in the author list, but on most papers that I see the lab head is listed as an author and is the corresponding author, taking responsibility for the programme of research that they have devised and continue to direct.

Jenny’s post appeared after we had finalised our guidelines, but we may consider that issue in a later revision, if it turns out that things are not as cosy here as I naively assume.